Lumen | Zenith

A JavaScript utility that provides text processing routines for post content.

I'm creating Lumen to lay the foundations of a social networking platform that I have in mind. And, as you can easily guess, users in a social network interact mainly through a component that allows them to publish text, links, images, emoji and much more. The library provides autolinking and extraction for URLs, usernames, lists, and hashtags.

Based on twitter-text.

NPM Users

Install it with: npm install @zenith-to/lumen-js

The lumen namespace is exported, making it available as such:

var lumen = require('@zenith-to/lumen-js')
lumen.autoLink(lumen.htmlEscape('#hello < @world >'))

Extraction Examples

// basic extraction
var usernames = lumen.extractMentions("Mentioning @vincent and @james")
// usernames == ["vincent", "james"]

Auto-linking Examples

lumen.autoLink("link @user, please #request");

lumen.autoLink("link @user, and expand url... http://z.to/0JG5Mcq", {
    urlEntities: [
        {
          "url": "http://z.to/0JG5Mcq",
          "display_url": "blog.zenith.to/2025/05/zenith…",
          "expanded_url": "http://blog.zenith.to/2025/05/zenith-is-online",
          "indices": [
            30,
            48
          ]
        }
    ]});

Content Parsing

lumen exposes a "parseText" method that will return the following fields:

• weightedLength: Integer that indicates the weighted length calculated by the algorithm above.
• permillage: Integer value corresponding to the ratio of consumed weighted length to the maximum weighted length.
• valid: Boolean indicating whether it is a valid content (for posts).
• dispayRangeStart: Integer with start index on the text string
• displayRangeEnd: Integer with end index on the text string (inclusive)
• validDisplayRangeStart: Integer indicating the valid start index on the text string
• validDisplayRangeEnd: Integer indicating the valid end index on the text string. This can be lesser than displayRangeEnd (inclusive).

var content = "This is a test post";
lumen.parseText(content);
/* Returns:
  {
    weightedLength: 20,
    permillage: 71,
    valid: true,
    displayRangeEnd: 19,
    displayRangeStart: 0,
    validRangeEnd: 19,
    validRangeStart: 0
  }
*/